Combining Machine Learning and Physics Modelling to Determine the Natural Cave Property with Fracturing Curves

Abstract

ABSTRACT Currently, it is challenging to determine the geological parameters of natural caves in deep reservoirs accurately. By inheriting the advantages of Machine Learning (ML) method and physics modelling, a novel ML-Physics method is developed to determine the geological parameters of natural caves based on the data obtained during Hydraulic Fracturing (HF) operation. The process of ML-Physics method is divided into preparation-stage and operation-stage. The preparation-stage happens before HF operation without the limitation of computational time, during which the implicit relationship between cave property and fracturing curve is generated. The computational time of the operation-stage is limited because the geological parameters of natural caves should be determined in real time during HF operation. During the operation-stage, the physical modelling based inverse analysis method is carried out, in which the initial value is chosen based on the results of preparation-stage. Results show that, with the same target error, the iteration step required by ML-Physics method is much less than that of traditional inverse analysis method. With the same iteration steps, the error of ML-Physics method is lower than that of the traditional inverse analysis method. The ML-Physics method is potentially useful to optimize the HF design in real time. INTRODUCTION The exploitation of oil and gas resources is very important for maintaining world energy security and keeping economic development. It is necessary to determine the geometry and geological parameters of natural caves to optimize the hydraulic fracturing design (Ma et al., 2018; Ali et al., 2019). Currently, seismic tomography is widely used to detect the distribution of natural fractures and caves (Chalikakis et al., 2011; Xu et al., 2016). With the rapid development of Internet of Things (IoT) technology, large quantities of monitoring data can be obtained in real time (Wang et al., 2022). Therefore, it is critical to develop an available method to determine the geological parameters according to the evolution of fracturing curves.